RESEARCH ARTICLE Open Access

Statin use and the risk of ovarian andendometrial cancers: a meta-analysisYizi Wang, Fang Ren, Zixuan Song, Peng Chen, Shuang Liu and Ling Ouyang*

Abstract

Background: The relationship between statin use and the risk of ovarian or endometrial cancer remains controversial.Here, we investigated the relationship between statin use and the risk of ovarian and endometrial cancers.

Methods: We conducted a meta-analysis using articles retrieved from the PubMed, Embase, and Web of Sciencedatabases. All original comparative studies published in English that were related to statin use and the risk of ovarian orendometrial cancer were included.

Results: This meta-analysis included 19 studies enrolling 1,999,362 female subjects and 19,849 cancer cases (7,948ovarian cancer cases and 11,901 endometrial cancer cases). The overall analysis indicated that statin use did notsignificantly reduce the risk of ovarian cancer [relative risk (RR) = 0.88, 95% confidence interval (CI) 0.76–1.03, p = 0.12] orthe risk of endometrial cancer (RR = 0.88, 95% CI 0.78–1.00, p = 0.05.) Subgroup analyses based on study type,percentage of cancer cases, study location, and quality of studies also supported our conclusions. No association wasfound between long-term statin use (> 5 years) and the risk of ovarian cancer (RR = 0.73, 95% CI 0.51–1.04, p = 0.08) orendometrial cancer (RR = 0.79, 95% CI 0.58–1.08, p = 0.14).

Conclusions: Statin use did not lower the risk of ovarian cancer or endometrial cancer. The long-term use ofstatins (> 5 years) was not associated with a reduction in the risk of ovarian or endometrial cancer.

Keywords: Statin, Ovarian cancer, Endometrial cancer

BackgroundOvarian and endometrial cancers are the two most com-monly diagnosed gynecological cancers, with the highestmortality of all gynecological cancers [1]. Ovarian canceris the leading cause of cancer death in gynecological ma-lignancies and the fifth leading cause of all cancers inwomen in the US [2]. Endometrial cancer is the mostcommon gynecological malignancy in the US, with adeath rate that has increased gradually [3].Esposito et al. reported in a meta-analysis that metabolic

syndrome is associated with an increased risk of ovarianand endometrial cancers [4]. Hydroxymethylglutaryl-CoAreductase inhibitors (statins) are cholesterol-loweringdrugs that are primarily used for hypercholesterolemiatherapy. They also have antitumor effects, which havebeen reported in experimental research on many cancers[5]. In recent years, the pharmacological functions of

statins were found to be effective in several femalereproductive diseases [6–9]. Statins inhibit 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA) reductase, thekey enzyme in the mevalonate pathway for cholesterolsynthesis. They can down regulate products of the meva-lonate pathway, such as farnesyl diphosphate and geranyldiphosphate, which may contribute to regulating tumorcell growth, motility, and differentiation [10, 11].Epidemiological studies of statins and the risk of endo-

metrial or ovarian cancers have shown inconsistent re-sults. A recent population-based study of 161,808postmenopausal women indicated that statin use canreduce the risk of endometrial cancer but not that of ovar-ian cancer [12]. In contrast, another study reported thatstatin use reduced the risk of ovarian cancer [13]. Previousstudies have reported that statin use was associated with areduced risk of ovarian or endometrial cancer [14, 15].Here, we conducted a meta-analysis to investigate the

relationship between statin use and the risk of endo-crine-related gynecological cancers.

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

* Correspondence: ouyl@sj-hospital.orgDepartment of Obstetrics and Gynecology, Shengjing Hospital of ChinaMedical University, Shenyang 110004, People’s Republic of China

Wang et al. BMC Cancer (2019) 19:730 https://doi.org/10.1186/s12885-019-5954-0

MethodsSearch strategyThis study was conducted in accordance with the PreferredReporting Items for Systematic Reviews and Meta-Analyses(PRISMA) guidelines (http://www.prisma-statement.org/).A comprehensive search was performed of the PubMed,Embase, and Cochrane databases to August 1, 2018. Thefollowing MeSH and main keywords were used: “hydroxy-methylglutaryl-CoA reductase inhibitor,” “ReductaseInhibitors, Hydroxymethylglutaryl-CoA,” “Inhibitors,HMG-CoA Reductase,” “Reductase Inhibitors, HMG-CoA,” “HMG-CoA Reductase Inhibitors,” “Statins, HMG-CoA,”“Statins,” “Hydroxymethylglutaryl-Coenzyme A Inhibitors,”and associated terms; and “Ovarian Neoplasm,” “OvaryNeoplasm,” “Ovary Cancer,” “Ovarian Cancer,” “Cancer ofOvary,” “Endometrial Neoplasm,” “Endometrial Carcinoma,” “Endometrial Cancer,” “Endometrium Cancer,”“Endometrium Carcinoma,” and associated terms. Thelanguage was restricted to English. For multiple-armcomparative studies, we extracted data only from thearms that matched our eligibility criteria. We alsoperformed manual searches of the reference lists ofthe selected studies to retrieve all relevant data.

Inclusion and exclusion criteriaStudies were selected according to the PICOS (popula-tion, intervention, comparison, outcomes, and study de-sign) guidelines if they met the following inclusioncriteria: (1) population: the entire population of womenrecruited to a certain trial, and the endpoint is incidenceof ovarian or endometrial cancer; any histologic types ofcancer were included for ovarian and endometrialcancers, including ovarian borderline malignancies; (2)intervention: patients used statins; (3) comparison: pa-tients used statin versus control (placebo or no statin);(4) outcomes: studies compared morbidity between twogroups using odds ratio (OR), relative risk (RR), or haz-ard ratio (HR) estimates with 95% confidence intervals(CIs), or provided sufficient data to compute morbidity;and (5) study design: studies were comparative [i.e., ran-domized control trials (RCTs) or observational studies].Exclusion criteria were as follows: (1) population:

mixed patients with other cancer types or patients withindeterminate cancer; (2) intervention: patients did notuse statins; (3) comparison: patients used statin versusother drugs; (4) outcomes: studies lacked morbidity data;and (5) study design: single-arm study without a controlgroup.

Data extraction and quality assessment of includedstudiesTwo reviewers (YW and FR) reviewed and assessed eachof the included studies. Data extraction was performedindependently. The other authors (ZS, PC, and SL)

reviewed the included studies as well for inclusion. Thefollowing information was extracted from the includedstudies: first author, publication year, country of study,study type, cancer site and cases, number of subjects,percentage of cancer cases, study period, cases of statinuse and control groups, follow-up time, and multivari-able adjusted RR estimates with corresponding 95% CIs.The Jadad scale was used to assess the quality of theRCTs [16], and studies with 3 or more points were iden-tified as high quality [17]. The Newcastle-Ottawa Scale(NOS) was used to assess the quality of the observationalstudies [18], and studies with 7 or more points weredeemed high quality. Any disagreements were resolved byconsensus between the two reviewers (YW and FR).

Statistical analysisThis meta-analysis used the RR to estimate morbidity.All analyses used a random-effects model. The statis-tical values are reported with 95% CIs, and a two-tailed p < 0.05 was considered statistically significant.Heterogeneity among studies was tested by Cochran’sQ test (reported with a χ2 value and p value, with ap < 0.1 considered statistically significant) and the I2

statistic [19, 20]. I2 with values of 25, 50, and 75%demonstrated low, moderate, and high levels of het-erogeneity, respectively [21]. We also conducted sen-sitivity analyses to assess the robustness of the results[22]. Subgroup analyses were conducted to exploresources of heterogeneity. Publication bias wasassessed using Begg’s and Egger’s tests [23, 24]. Allanalyses were performed using Stata software version12.0 (2011; Stata Corp., College Station, TX, USA).

ResultsCharacteristics of selected studies

A flow chart of the search strategy is illustrated inFig. 1. The search retrieved 2,203 published studies.After removing duplicates, screening the title andabstract, and further evaluation, a total of 20 studieswere included in our meta-analysis. Among these,one publication was excluded because it providedduplicated data [25]. The remaining 19 studiesincluded two RCTs and 17 observational studies[12–15, 26–40]. Nine studies enrolled endometrialcancer patients, three enrolled ovarian cancer pa-tients, and seven enrolled both. The cancer patientswere from the population of women studied. Thestudies were carried out in eight countries, which in-cluded North America [12, 13, 26, 27, 29, 30, 32, 34,39, 40], Europe [14, 15, 28, 31, 33, 35, 36, 38], andAsia [37]. Detailed information from the includedstudies is presented in Table 1.

Wang et al. BMC Cancer (2019) 19:730 Page 2 of 10

Risk of ovarian cancer among statin usersThe meta-analysis included 10 studies of ovarian can-cer. Our pooled analysis indicated that statin use didnot reduce the risk of ovarian cancer (RR = 0.88, 95%CI = 0.76–1.03, p = 0.12) (Fig. 2). A moderately signifi-cant heterogeneity was observed among the studies(χ2 = 20.00, p = 0.02, I2 = 55.0%). Sensitivity analyseswere carried out by excluding studies one-by-one, andwe found that the corresponding pooled RRs werenot significantly altered, which affirmed the robust-ness of the result (Additional file 1: Figure S1). Weconducted subgroup analyses of statin use and therisk of ovarian cancer based on study type, percentageof cancer cases, geographical study location (contin-ent), and quality of studies. There was no significantassociation between statin use and the risk of ovariancancer in the RCTs (RR = 0.20, 95% CI = 0.01–4.07,p = 0.30), cohort studies (RR = 0.98, 95% CI = 0.77–1.24, p = 0.85), or case-control studies (RR = 0.82, 95%CI = 0.65–1.04, p = 0.10). Subgroup analysis based onthe percentage of cancer cases indicated that therewas no significant association in groups with ≥1%(RR = 0.76, 95% CI = 0.55–1.05, p = 0.10) or < 1%(RR = 0.97, 95% CI = 0.83–1.14, p = 0.74). Subgroupanalysis of study location showed no reduced risk ofovarian cancer among statin users in North America(RR =0.88, 95% CI = 0.69–1.12, p = 0.29) or Europe(RR =0.90, 95% CI = 0.72–1.12, p = 0.34). Addition-ally, no decreased risk of ovarian cancer in statin

users was observed, whether studies were high qual-ity or low quality (Table 2). There was no evidenceof publication bias (Begg’s test, p = 0.86, [Additional file 2:Figure S2A]; Egger’s test, p = 0.34, [Additional file 2:Figure S2B]).

Risk of endometrial cancer among statin usersThe meta-analysis included 16 studies of endometrialcancer. There was no significant association betweenthe use of statins and the risk of endometrial cancer(RR = 0.88, 95% CI = 0.78–1.00, p = 0.05) (Fig. 3). Amoderately significant heterogeneity was observedamong the studies (χ2 = 47.50, P = 0.00, I2 = 66.3%).Sensitivity analyses were carried out by excludingstudies one-by-one, and we found that six studies [15,26, 30, 32, 34, 35] seemed to affect the heterogeneity.Excluding any one of these studies from the overallanalysis resulted in statistically significant results, butthe upper limits of the 95% confidence interval wereapproximately 1.00 (Additional file 3: Figure S3). Wealso conducted subgroup analyses by study type, per-centage of cancer cases, study location, and quality ofstudies. There was no significant association betweenstatin use and the risk of endometrial cancer in theRCTs (RR = 0.72, 95% CI = 0.19–2.68, p = 0.62), cohortstudies (RR = 0.93, 95% CI = 0.80–1.10, p = 0.38), orcase-control studies (RR = 0.80, 95% CI = 0.62–1.03,p = 0.09). Subgroup analysis based on the percentageof cancer cases indicated that there was no significant

Fig. 1 Flow chart of our systematic review and meta-analysis

Wang et al. BMC Cancer (2019) 19:730 Page 3 of 10

Table

1Baselinecharacteristicsof

theinclud

edstud

ies

Stud

yCou

ntry

Stud

ype

riod

Stud

ytype

All

female

objects

ECOC

ECOC

Follow

up (med

ian/

mean

years)

NOS

percen

tage

ofcancer

cases

User

Non

user

percen

tage

ofcancer

cases

User

Non

user

RR95%

CI

RR95%

CI

Sperling2017

Den

mark

2000–2009

Case-control

77509

6.94%

610

4772

NA

NA

NA

1.03

0.94–1.14

NA

NA

NA

7

Urpilainen

2018

Finland

1996–2011

Case-control

748282

NA

NA

NA

0.04%

159

144

NA

NA

0.99

0.78–1.25

5.4

7

Jacobs

2011

USA

1997–2007

Coh

ort

73196

0.63%

134

327

NA

NA

NA

>5y:0.65

<5y:1.11

0.45–0.94

0.84–1.46

NA

NA

NA

8

Friedm

an2008

USA

1994–2003

Coh

ort

169261

0.12%

199

NA

0.05%

79–

1.13

0.98–1.31

0.83

0.66–1.05

4.9

7

Kabat2018

USA

1993–1998

Coh

ort

24208

0.48%

13102

0.48%

15100

0.96

0.54–1.7

1.24

0.72–2.12

NA

6

Leun

g2013

China

2000–2008

Case-control

18055

1.23%

13209

NA

NA

NA

0.43

0.20–0.93

NA

NA

4.2

7

Kaye

2004

UK

1990–2002

Case-control

8978

0.81%

370

1.01%

685

0.50

0.10–1.90

1.00

0.40–2.70

6.4

8

Haukka2010

Finland

1996–2005

Coh

ort

473302

0.36%

836

885

NA

NA

NA

1.05

0.95–1.15

NA

NA

37

Fortun

y2009

USA

2001–2005

Case-control

936

50.11%

114

355

NA

NA

NA

1.30

0.81–2.10

NA

NA

NA

7

Baandrup

2015

Den

mark

2000–2011

Case-control

62809

NA

NA

NA

6.53%

434

3669

NA

NA

0.98

0.87–1.10

NA

8

Blais2000

Canada

1988–1994

Case-control

4029

7.10%

26260

NA

NA

NA

0.30

0.11–0.80

NA

NA

2.7

5

Clearfield

2001

USA

NA

RCT

997

0.40%

13

0.20%

02

0.33

0.03–3.19

0.20

0.01–4.15

5.2

6a

Lavie2013

Israel

2003–2010

Case-control

682

31.52%

70145

18.48%

3888

0.55

0.37–0.81

0.49

0.28–0.81

NA

8

Yu2009

USA

1990–2004

Coh

ort

73336

0.77%

18550

0.44%

12314

0.67

0.39–1.17

0.69

0.32–1.49

5.6

8

Coo

gan2007

USA

1991–2005

Case-control

4641

4.59%

19194

NA

NA

NA

1.30

0.70–2.40

NA

NA

NA

5

Desai2018

USA

1993–1998

Coh

ort

161808

0.85%

781299

0.47%

62701

0.74

0.59–0.94

1.15

0.89–1.50

10.8

8

Arim

a2017

Finland

1996–2011

Case-control

92366

0.64%

270

320

NA

NA

NA

0.78

0.65–0.94

NA

NA

5.5

7

Strand

berg

2004

Nordiccoun

tries

1988–1999

RCT

827

0.73%

33

NA

NA

NA

1.03

0.21–5.14

NA

NA

10.4

5a

Akinw

unmi2018

USA

1992–2008

Case-control

4140

NA

NA

NA

49.28%

188

1852

NA

NA

0.68

0.54–0.85

NA

8

RCTrand

omized

controlledtrial,NOSNew

castle–O

ttaw

aScale,

NAno

tap

plicab

lea Jad

adscalewas

used

toassess

thequ

ality

oftherand

omized

clinical

trials

Wang et al. BMC Cancer (2019) 19:730 Page 4 of 10

association in groups with ≥1% (RR = 0.79, 95% CI =0.54–1.15, p = 0.21) or < 1% (RR = 0.90, 95% CI =0.78–1.04, p = 0.14) of cancer cases in the studypopulation. Subgroup analysis of studies based ongeographical location showed a significantly reducedrisk of endometrial cancer among statin users inAsia (RR = 0.43, 95% CI = 0.20–0.94, p = 0.03), butnone in North America (RR = 0.90, 95% CI = 0.73–

1.11, p = 0.31) or Europe (RR = 0.89, 95% CI = 0.75–1.05, p = 0.18). Again, no significant association be-tween statin use and risk of endometrial cancer wasobserved in studies of high quality or low quality(Table 2). There was evidence of potential publica-tion bias (Begg’s test, P = 0.434, [Additional file 4:Figure S4A]; Egger’s test, p = 0.03, [Additional file 4:Figure S4B]).

Fig. 2 Overall analysis of statin use and the risk of ovarian cancer

Table 2 Subgroup analyses of statin use and risk of endocrine-related gynecologic cancers

Characteristics Ovarian Cancer Endometrial Cancer

Study number RR(95%CI) P value Heterogeneity Study number RR(95% CI) P value Heterogeneity

Study type 10 0.88 (0.76,1.03) 0.12 55.0% 16 0.88 (0.78,1.00) 0.05 66.3%

RCT 1 0.20 (0.01,4.07) 0.30 0.0% 2 0.72 (0.19,2.68) 0.62 0.0%

Cohort 4 0.98 (0.77,1.24) 0.85 38.9% 7 0.93 (0.80,1.10) 0.38 66.5%

Case-control 5 0.82 (0.65,1.04) 0.10 70.6% 7 0.80 (0.62,1.03) 0.09 74.1%

Percentage of cancer cases

≥ 1% 4 0.76 (0.55,1.05) 0.10 76.8% 6 0.79 (0.54,1.15) 0.21 76.5%

< 1% 6 0.97 (0.83,1.14) 0.74 16.6% 10 0.90 (0.78,1.04) 0.14 61.8%

Study location

North America 6 0.88 (0.69,1.12) 0.29 58.1% 9 0.90 (0.73,1.11) 0.31 64.2%

Europe 4 0.90 (0.72,1.12) 0.34 52.7% 6 0.89 (0.75,1.05) 0.18 72.2%

Asia 0 – – – 1 0.43 (0.20,0.94) 0.03 0.0%

Quality of studies

High 9 0.86 (0.74,1.01) 0.07 57.4% 13 0.88 (0.78,1.01) 0.06 68.5%

Low 1 1.24 (0.72,2.12) 0.44 0.0% 3 0.80 (0.39,1.62) 0.53 67.4%

Dduration of statin use

long-term use 5 0.73 (0.51,1.04) 0.08 58.9% 6 0.79 (0.58,1.08) 0.14 58.2%

Study in diabetics 1 0.96 (0.75,1.23) 0.75 0.0% 1 0.78 (0.65,0.94) 0.01 0.0%

Wang et al. BMC Cancer (2019) 19:730 Page 5 of 10

Long-term statin use and the risk of ovarian orendometrial cancerThe meta-analysis included five studies of ovarian cancer[13, 14, 30, 36, 39] and six studies of endometrial cancer[15, 30, 33, 34, 36, 39] that analyzed long-term statin use(> 5 years). There was no association between long-termstatin use (> 5 years) and the risk of endometrial cancer(RR = 0.79, 95% CI = 0.58–1.08, p = 0.14) or ovarian can-cer (RR = 0.73, 95% CI = 0.51–1.04, p = 0.08) (Fig. 4).

DiscussionThe association between statin use and the risk of endo-metrial or ovarian cancer remains controversial. Sperlinget al. [15] reported no association between statin useand the risk of endometrial cancer, whereas Desai et al.[12] reported that statins reduced the risk of endometrialcancer. Some studies have reported a significant associ-ation between statin use and the risk of ovarian cancer[13, 25, 36], but others found no such association [12,31]. In 2014, Liu et al. [41] performed a meta-analysis ofthe effect of statin use on the risk of gynecologicalcancers and reported that statin use lowered the risk ofovarian cancer but not that of endometrial cancer. Thedebate has continued after that publication, and the re-sults of more recent studies have also been inconsistent.Our meta-analysis included 19 studies enrolling 1,999,362

women. Seven of these studies were published after the lastmeta-analysis in 2014 and included an additional 1,171,122subjects. Our pooled analysis indicated that statin use

did not reduce the risk of ovarian or endometrialcancer. In addition, we conducted subgroup analysesby study type, percentage of cancer cases, quality ofstudies, and study location to explore sources of het-erogeneity. Two RCTs, six cohort studies and 11case-control studies were included, and there was nosignificant association between the use of statins andthe risk of ovarian or endometrial cancer based onstudy type. We used the Jadad scale and NOS criter-ion to assess the quality of the RCTs and observa-tional studies, respectively. The two RCTs weredeemed to be of high quality, with Jadad total scores≥3, and 14 observational studies had NOS scores ≥7,identifying them as high quality. Subgroup analysesbased on the quality of studies indicated that no sig-nificant association between statin use and the risk ofendometrial cancer was observed in studies of highquality or low quality. The studies we included haddifferent numbers of participants, most included largepopulations, and the percentage of cancer patientswas low and varied. However, we found no significantdifferences in the association of statin use and therisk of ovarian or endometrial cancer in groups withpercentages of cancer cases ≥1% versus groups with< 1%. All of the above subgroup analyses supportedour pooled conclusions. In the subgroup analysesbased on study location, our results showed a similartendency, with no reduced risk of ovarian or endo-metrial cancer in statin users, although a subgroup of

Fig. 3 Overall analysis of statin use and the risk of endometrial cancer

Wang et al. BMC Cancer (2019) 19:730 Page 6 of 10

statin users in Asia did have a significantly reduced risk ofendometrial cancer. This result might be due to regionaldisparities or racial differences. People in different coun-tries have different dietary habits. Patel et al. reported thatmetabolic syndrome was less prevalent in Asian popula-tions than it is in US populations [42]. Therefore, the ef-fects of statin use may vary according to location.However, this meta-analysis subgroup (endometrial cancerpatients in Asia) included only one study conducted inAsia, which enrolled fewer subjects than in other studies.Hence, the effects of statin use in endometrial cancer pa-tients require further exploration, and our results requirevalidation from additional RCTs enrolling large populations.Ovarian and endometrial cancers may be associated

with diabetes mellitus. Previous studies reported thatpatients with diabetes mellitus had an increased risk of

ovarian cancer, and patients with ovarian cancer and dia-betes had poor survival [43, 44]. The incidence of endo-metrial cancer is increasing due to global increases indiabetes and obesity [45, 46], suggesting that we shouldtarget women with diabetes. We planned to conductsubgroup analyses of studies enrolling women with dia-betes, but the number of studies was limited. Althoughthe risk of endometrial cancer was significantly reducedin statin users, the results require validation with add-itional studies enrolling larger populations.Studies investigating the effects of long-term statin use

on ovarian or endometrial cancer have reported incon-sistent results. Some studies found no association [34].In contrast, the meta-analysis of Liu et al. [41] reportedthat long-term statin use significantly reduced the risk ofovarian cancer. Our subgroup analyses of long-term

Fig. 4 Subgroup analyses of long-term statin use and the risk of ovarian and endometrial cancer. a Ovarian cancer; b Endometrial cancer

Wang et al. BMC Cancer (2019) 19:730 Page 7 of 10

statin use included three more studies and 5,817 add-itional subjects than the previous meta-analysis [41] andshowed no significant association between long-termstatin use and the risk of ovarian or endometrial cancer.As we found no reduction in the risk of ovarian or endo-metrial cancer in statin users, long-term statin use alsohad no protective effect. The majority of studies of theeffect of long-term statin use indicated no reduced riskof ovarian or endometrial cancer in statin users. Indeed,a recent observational study by Desai et al. [12] reportedthat the use of pravastatin was associated with an in-creased risk of ovarian cancer. This study enrolled alarge number of female subjects (161,808) and investi-gated the relationship between different drug types andcancer risk. Thus, the long-term effect of statins muststill be considered unanswered, and more RCTs or high-quality observational studies are needed to verify the effectsof long-term statin use on ovarian or endometrial cancer.Our study has some advantages. The meta-analysis

included 19 studies enrolling 1,999,362 female sub-jects and 19,849 cancer cases. Our inclusion and ex-clusion criteria were based on PICOS criteria. TheJadad scale was used to assess the quality of theRCTs, and the NOS criterion was used to evaluatethe quality of the observational studies; most of thestudies were high quality. The 2014 meta-analysis [41]investigated the association between statin use andthe risk of gynecological cancers. In contrast, we in-vestigated the association between statin use and therisk of ovarian and endometrial cancers separately.Our meta-analysis included seven more studies and alarger number of participants than the previous meta-analysis, and our results differed from those of theearlier study. Our subgroup analyses based on studytype, percentage of cancer cases, study location, andquality of studies largely supported our conclusions.There were some limitations in our meta-analysis,

which will be resolved in future studies. First, in theselected studies, women were prescribed statins pri-marily to treat metabolic syndrome, which may inde-pendently induce ovarian or endometrial cancer. As aresult, it was not possible to pool metabolic status forthe observational studies included in this meta-ana-lysis. Moreover, in some studies, patients were pre-scribed other medications in addition to statins. Asother medications may be associated with the risk ofcancer, for example, the use of metformin or oralcontraceptives may lower the risk of gynecologic can-cer [47, 48], and hormone replacement therapy mayinfluence the risk of gynecologic cancer [49], therewas the possibility that our results had bias, whichcould be of concern. Hence, to confirm the conclu-sions of this meta-analysis, further RCTs enrolling lar-ger populations to investigate the association between

statin use and the risk of ovarian or endometrial can-cer are needed.Second, the heterogeneity of our study was significant.

Our study included 2 RCTs and 17 observational studies;the two RCTs included 12 total cancer cases, whereasthe largest studies included approximately 5,000 cancercases [15]. Among the included observational studies,the study by Lavie et al. [36] is a case-control trial inwhich controls were individually matched to cases ac-cording to age, sex, clinic and ethnic group. Therefore,the study excluded many confounding factors, whichmight lead to a more reliable result. As it has been re-ported that statin use could reduce the risk of endomet-rial and ovarian cancers, we cannot ignore the importantfinding. Although our pooled analysis was negative, themajority of the studies we included were retrospective,which may contain selection, information, and con-founding bias. More RCTs or high-quality observationalstudies are needed to confirm or update our meta-ana-lysis. In addition, we extracted RR data directly or indir-ectly to represent the relationship between statin useand ovarian or endometrial cancer. However, each of theincluded studies utilized different study designs accord-ing to the different statins or different statin doses. Moststudies observed the outcomes of overall statin use(rather than that of a particular drug) and cancer risk.Other studies classified statins as lipophilic or hydro-philic [13, 37] or separately observed different drugs[15, 33, 35, 39]. As different statins have differenthalf-lives, different types or doses of statin have been pro-posed to contribute differently to clinical outcomes. How-ever, the results could not be combined because of suchdifferences in the included studies. Most studies did notrecord the duration of statin use, and other studies evalu-ating statin duration were not coincident in the observa-tion time. Thus, we could not combine the resultsaccording to the duration of statin use specifically, butonly extracted data for long-term statin use (> 5 years).Moreover, the studies we included differed in terms of re-search methods, follow-up time, and source of the popula-tion; any of these factors may have induced studyheterogeneity.Third, the sensitivity analysis indicated that excluding

studies one-by-one did not alter the correspondingpooled RRs in the ovarian cancer group, which affirmedthe robustness of the result. However, for the studies ofendometrial cancer, we found that six studies seemed toaffect the heterogeneity. When we excluded any one ofthese six studies from the overall analysis, statisticallysignificant results were obtained. However, the upperlimits of the 95% confidence intervals were approxi-mately 1.00. Thus, as we did not think that excluding somany studies was appropriate, we maintained the ori-ginal pooled results. In addition, when we conducted

Wang et al. BMC Cancer (2019) 19:730 Page 8 of 10

analyses of statin use and the risk of ovarian cancer,there was no evidence of publication bias. However, inthe overall analyses of endometrial cancer, althoughthere was no significance in Begg’s test, Egger’s test wasstatistically significant, which indicated a potential publi-cation bias. The most likely explanation is the preferen-tial publication of positive findings.Fourth, the number of RCTs investigating statin use

and the risk of ovarian or endometrial cancer was lim-ited. The search retrieved two RCTs with 12 cancercases but 17 observational studies with 19,837 cancercases. Moreover, in these RCTs, statins were used totreat patients with coronary heart disease or hyperlip-emia. Cancer was not a primary outcome. Hence, the re-sults from the two RCTs were not powerful enough toconclude the outcome of cancers. Thus, further RCTsinvestigating statin use and the risk of ovarian or endo-metrial cancer will be required to confirm these results.

ConclusionsThis meta-analysis identified no reduction in the risk ofovarian or endometrial cancer for patients who used sta-tins. In addition, the long-term use of statins (> 5 years)was not associated with a reduction in the risk of ovar-ian or endometrial cancer. To confirm the conclusionsof this meta-analysis, further RCTs enrolling larger pop-ulations to investigate the association between statin useand the risk of ovarian or endometrial cancer areneeded.

Additional files

Additional file 1: Figure S1. Sensitivity analyses based on statin useand the risk of ovarian cancer. (TIF 1190 kb)

Additional file 2: Figure S2. Publication bias in terms of statin use andthe risk of ovarian cancer. (A) Begg’s test; (B) Egger’s test. (TIF 1616 kb)

Additional file 3: Figure S3. Sensitivity analyses based on statin useand the risk of endometrial cancer. (TIF 1344 kb)

Additional file 4: Figure S4. Publication bias in terms of statin use andthe risk of endometrial cancer. (A) Begg’s test; (B) Egger’s test.(TIF 1658 kb)

AbbreviationsCI: Confidence interval; HMG CoA: 3-hydroxy-3-methylglutaryl coenzyme A;HR: Hazard ratio; NA: Not applicable; NOS: Newcastle-Ottawa Scale; OR: Oddsratio; PICOS: Population, intervention, comparison, outcomes, and studydesign; PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses; RCT: Randomized control trial; RR: Relative risk;Statins: Hydroxymethylglutaryl-CoA reductase inhibitors

AcknowledgementsNot applicable.

Authors’ contributionsLO and YW conceived the study idea and devised the study methodology.FR and ZS participated in the design and coordination of the study. YW andFR screened identified literature, conducted data extraction, analysed thenarrative review findings and conducted the meta-analysis. PC and SLprovided specific support in quantitative data analysis. YW wrote the

manuscript. All authors contributed to improving the manuscript, read andapproved the version of the manuscript to be published. All authors takeresponsibility for appropriate content.

FundingThe project was supported by National Natural Science Foundation of China(Grant number 81501235). The funding body had no role in the design ofthe study, collection, analysis, or interpretation of the data, or in the writingof the manuscript.

Availability of data and materialsInput data for the analyses are available from the corresponding author onrequest.

Ethics approval and consent to participateNot applicable.

Consent for publicationNot applicable.

Competing interestsThe authors declare that they have no competing interests.

Received: 24 October 2018 Accepted: 18 July 2019

References1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2017. CA Cancer J Clin. 2017;

67(1):7–30.2. Arend RC, Londono-Joshi AI, Straughn JM Jr, Buchsbaum DJ. The Wnt/

beta-catenin pathway in ovarian cancer: a review. Gynecol Oncol. 2013;131(3):772–9.

3. Sorosky JI. Endometrial cancer. Obstet Gynecol. 2012;120(2 Pt 1):383–97.4. Esposito K, Chiodini P, Colao A, Lenzi A, Giugliano D. Metabolic syndrome

and risk of cancer: a systematic review and meta-analysis. Diabetes Care.2012;35(11):2402–11.

5. Iannelli F, Lombardi R, Milone MR, Pucci B, De Rienzo S, Budillon A, BruzzeseF. Targeting mevalonate pathway in Cancer treatment: repurposing ofstatins. Recent Pat Anticancer Drug Discov. 2018;13(2):184–200.

6. Kobayashi Y, Kashima H, Rahmanto YS, Banno K, Yu Y, Matoba Y, WatanabeK, Iijima M, Takeda T, Kunitomi H, et al. Drug repositioning of mevalonatepathway inhibitors as antitumor agents for ovarian cancer. Oncotarget.2017;8(42):72147–56.

7. Zeybek B, Costantine M, Kilic GS, Borahay MA. Therapeutic roles ofstatins in gynecology and obstetrics: the current evidence. Reprod Sci.2018;25(6):802–17.

8. Schointuch MN, Gilliam TP, Stine JE, Han X, Zhou C, Gehrig PA, Kim K, Bae-Jump VL. Simvastatin, an HMG-CoA reductase inhibitor, exhibits anti-metastatic and anti-tumorigenic effects in endometrial cancer. GynecolOncol. 2014;134(2):346–55.

9. Kato S, Liberona MF, Cerda-Infante J, Sanchez M, Henriquez JF, Bizama C,Bravo ML, Gonzalez P, Gejman R, Branes J, et al. Simvastatin interferes withcancer 'stem-cell' plasticity reducing metastasis in ovarian cancer. EndocrRelat Cancer. 2018. https://doi.org/10.1530/erc-18-0132.

10. Laezza C, Malfitano AM, Proto MC, Esposito I, Gazzerro P, Formisano P,Pisanti S, Santoro A, Caruso MG, Bifulco M. Inhibition of 3-hydroxy-3-methylglutaryl-coenzyme a reductase activity and of Ras farnesylationmediate antitumor effects of anandamide in human breast cancer cells.Endocr Relat Cancer. 2010;17(2):495–503.

11. Cuello FM, Kato CS, Diaz SD, Owen G. Effects of statins in cancer. Rev MedChil. 2013;141(2):227–36.

12. Desai P, Wallace R, Anderson ML, Howard BV, Ray RM, Wu C, Safford M,Martin LW, Rohan T, Manson JE, et al. An analysis of the associationbetween statin use and risk of endometrial and ovarian cancers in theWomen's Health Initiative. Gynecol Oncol. 2018;148(3):540–6.

13. Akinwunmi B, Vitonis AF, Titus L, Terry KL, Cramer DW. Statin therapy andassociation with ovarian Cancer risk in the New England case control (NEC)study. Int J Cancer. 2018. https://doi.org/10.1002/ijc.31758.

14. Baandrup L, Dehlendorff C, Friis S, Olsen JH, Kjaer SK. Statin use and risk forovarian cancer: a Danish nationwide case-control study. Br J Cancer. 2015;112(1):157–61.

Wang et al. BMC Cancer (2019) 19:730 Page 9 of 10

15. Sperling CD, Verdoodt F, Friis S, Dehlendorff C, Kjaer SK. Statin use and riskof endometrial cancer: a nationwide registry-based case-control study.Biomed Res Int. 2017;96(2):144–9.

16. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ,McQuay HJ. Assessing the quality of reports of randomized clinical trials: isblinding necessary? Control Clin Trials. 1996;17(1):1–12.

17. Clark HD, Wells GA, Huet C, McAlister FA, Salmi LR, Fergusson D, Laupacis A.Assessing the quality of randomized trials: reliability of the Jadad scale.Control Clin Trials. 1999;20(5):448–52.

18. Stang A. Critical evaluation of the Newcastle-Ottawa scale for theassessment of the quality of nonrandomized studies in meta-analyses. Eur JEpidemiol. 2010;25(9):603–5.

19. Laird NM, Mosteller F. Some statistical methods for combining experimentalresults. Int J Technol Assess Health Care. 1990;6(1):5–30.

20. Dickersin K, Berlin JA. Meta-analysis: state-of-the-science. Epidemiol Rev.1992;14:154–76.

21. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency inmeta-analyses. Bmj. 2003;327(7414):557–60.

22. Copas J, Shi JQ. Meta-analysis, funnel plots and sensitivity analysis.Biostatistics. 2000;1(3):247–62.

23. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysisdetected by a simple, graphical test. Bmj. 1997;315(7109):629–34.

24. Begg CB, Mazumdar M. Operating characteristics of a rank correlation testfor publication bias. Biometrics. 1994;50(4):1088–101.

25. Rennert G, Rennert HS, Pinchev M, Lavie O. The effect ofbisphosphonates on the risk of endometrial and ovarian malignancies.Gynecol Oncol. 2014;133(2):309–13.

26. Coogan PF, Rosenberg L, Strom BL. Statin use and the risk of 10 cancers.Epidemiology. 2007;18(2):213–9.

27. Yu O, Boudreau DM, Buist DS, Miglioretti DL. Statin use and femalereproductive organ cancer risk in a large population-based setting. CancerCauses Control. 2009;20(5):609–16.

28. Kaye JA, Jick H. Statin use and cancer risk in the general practice researchdatabase. Br J Cancer. 2004;90(3):635–7.

29. Kabat GC, Kim MY, Chlebowski RT, Vitolins MZ, Wassertheil-Smoller S, RohanTE. Serum lipids and risk of obesity-related cancers in postmenopausalwomen. Cancer Causes Control. 2018;29(1):13–24.

30. Friedman GD, Flick ED, Udaltsova N, Chan J, Quesenberry CP Jr, Habel LA.Screening statins for possible carcinogenic risk: up to 9 years of follow-up of361,859 recipients. Pharmacoepidemiol Drug Saf. 2008;17(1):27–36.

31. Urpilainen E, Marttila M, Hautakoski A, Arffman M, Sund R, Ilanne-Parikka P,Arima R, Kangaskokko J, Puistola U, Laara E, et al. The role of metformin andstatins in the incidence of epithelial ovarian cancer in type 2 diabetes: acohort and nested case-control study. Bjog. 2018;125(8):1001–8.

32. Fortuny J, Sima C, Bayuga S, Wilcox H, Pulick K, Faulkner S, Zauber AG,Olson SH. Risk of endometrial cancer in relation to medical conditions andmedication use. Cancer Epidemiol Biomark Prev. 2009;18(5):1448–56.

33. Strandberg TE, Pyorala K, Cook TJ, Wilhelmsen L, Faergeman O,Thorgeirsson G, Pedersen TR, Kjekshus J. Mortality and incidence of cancerduring 10-year follow-up of the Scandinavian simvastatin survival study (4S).Lancet. 2004;364(9436):771–7.

34. Jacobs EJ, Newton CC, Thun MJ, Gapstur SM. Long-term use of cholesterol-lowering drugs and cancer incidence in a large United States cohort.Cancer Res. 2011;71(5):1763–71.

35. Haukka J, Sankila R, Klaukka T, Lonnqvist J, Niskanen L, Tanskanen A,Wahlbeck K, Tiihonen J. Incidence of cancer and statin usage--recordlinkage study. Int J Cancer. 2010;126(1):279–84.

36. Lavie O, Pinchev M, Rennert HS, Segev Y, Rennert G. The effect of statins on riskand survival of gynecological malignancies. Gynecol Oncol. 2013;130(3):615–9.

37. Leung HW, Chan AL, Lo D, Leung JH, Chen HL. Common cancer risk andstatins: a population-based case-control study in a Chinese population.Expert Opin Drug Saf. 2013;12(1):19–27.

38. Arima R, Marttila M, Hautakoski A, Arffman M, Sund R, Ilanne-Parikka P,Kangaskokko J, Laara E, Puistola U, Hinkula M. Antidiabetic medication,statins and the risk of endometrioid endometrial cancer in patients withtype 2 diabetes. Gynecol Oncol. 2017;146(3):636–41.

39. Clearfield M, Downs JR, Weis S, Whitney EJ, Kruyer W, Shapiro DR, Stein EA,Langendorfer A, Beere PA, Gotto AM. Air force/Texas coronaryatherosclerosis prevention study (AFCAPS/TexCAPS): efficacy and tolerabilityof long-term treatment with lovastatin in women. J Womens Health GendBased Med. 2001;10(10):971–81.

40. Blais L, Desgagne A, LeLorier J. 3-Hydroxy-3-methylglutaryl coenzyme areductase inhibitors and the risk of cancer: a nested case-control study.Arch Intern Med. 2000;160(15):2363–8.

41. Liu Y, Qin A, Li T, Qin X, Li S. Effect of statin on risk of gynecologic cancers:a meta-analysis of observational studies and randomized controlled trials.Gynecol Oncol. 2014;133(3):647–55.

42. Patel A, Huang KC, Janus ED, Gill T, Neal B, Suriyawongpaisal P, Wong E,Woodward M, Stolk RP. Is a single definition of the metabolic syndromeappropriate?--a comparative study of the USA and Asia. Atherosclerosis.2006;184(1):225–32.

43. Lee JY, Jeon I, Kim JW, Song YS, Yoon JM, Park SM. Diabetes mellitus andovarian cancer risk: a systematic review and meta-analysis of observationalstudies. Int J Gynecol Cancer. 2013;23(3):402–12.

44. Shah MM, Erickson BK, Matin T, McGwin G Jr, Martin JY, Daily LB, Pasko D,Haygood CW, Fauci JM, Leath CA 3rd. Diabetes mellitus and ovarian cancer:more complex than just increasing risk. Gynecol Oncol. 2014;135(2):273–7.

45. Jenabi E, Poorolajal J. The effect of body mass index on endometrial cancer:a meta-analysis. Public Health. 2015;129(7):872–80.

46. Pearson-Stuttard J, Zhou B, Kontis V, Bentham J, Gunter MJ, Ezzati M.Worldwide burden of cancer attributable to diabetes and high body-mass index: a comparative risk assessment. Lancet Diabetes Endocrinol.2018;6(6):e6–e15.

47. Wen Q, Zhao Z, Wen J, Zhou J, Wu J, Lei S, Miao Y. The associationbetween metformin therapy and risk of gynecological cancer in patients:two meta-analyses. Eur J Obstet Gynecol Reprod Biol. 2019;237:33–41.

48. Cho MK. Use of combined Oral contraceptives in Perimenopausal women.Chonnam Med J. 2018;54(3):153–8.

49. Sjogren LL, Morch LS, Lokkegaard E. Hormone replacement therapy and therisk of endometrial cancer: a systematic review. Maturitas. 2016;91:25–35.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Wang et al. BMC Cancer (2019) 19:730 Page 10 of 10